If you’re using trigs placed in grid mode, then I don’t think it matters.
1 Like
This concerns trigs (notes) live recording, not audio recording.
1 Like
bpm_trig_table.xlsx (13.9 KB)
The attachment has all the combinations of BPM and trig number (RLEN) that should give clickless loops. There’s a pretty simple equation that governs the relationship between sample rate, sample buffer size, and number of trigs:
mod(60*44100xRLEN/(4xBPM),64) = 0
I’m having a lot of fun with the 78.1 BPM and 20 trig long sample buffer.
3 Likes
Which is 78.125 bpm actually.
Which is meaningless to me if my set is 128bpm and my patterns are either 2, 4, or 8 bars long and Im sampling one bar loops.
Still, Im not getting any clicks that I can hear, so it doesnt bother me personally.
1 Like
Yes. All of the calculations were done at the x/24 actual BPM values. Spreadsheet has Octatrack’s displayed value.
1 Like
ABSOLUTELY ESSENTIAL TO KNOW if you prepare a loop for Octatrack in a DAW at exact tempo…
1 Like
Morning exploration was with two recording buffers. BPM set to 153.1. T1 was RLEN 64, T2 was RLEN 62. One track panned hard right, one hard left. Both record buffers are left open throughout.
I added a summary to the first post of the thread. Hopefully future folks who want to use the OT for sound-on-sound looping will find it useful. I certainly would have! Though I learned a lot about the OT along the way.
2 Likes
An update on this:
I’ve been using this to create clickless overdubbing loops with no issues. Any of the BPM/length combinations work for this. Really great for my needs.
However, I did notice that I get the occasional click when I stop recording. That’s not great if you’re hoping to save the loop you built up. A workaround is to record the loop to a second recording buffer and save that. There might be other ways.
2 Likes
Did Elektron ever acknowledge this? Let alone commit to fixing it?
Might be strange to read, there was obviously good reason to invent new machines that do it slightly different. There might be technical limitation that prevents any fix without sacrificing the machine’s intent. So the birth of those other machines is likely the very answer, just that MK2 is a republishing rather then a reinvention and as such has the same limitation.
They give the workaround of using FIN/FOUT to smooth things, or using ATT on playback. Those are genuinely fine for many applications.
My use case was really trying to push the record buffers to do seamless loops. I do a lot of sound-on-sound looping and having clicks sneak in ruined the experience.
I also want, for example, to be able to start playing back a loop at any point in it and have it loop while never hearing where the loop point in the original recording was. I’ll often record a loop and then realize I’ve started hearing the downbeat in a new place. I didn’t want there to be a little glitch (either a click or a cross fade) where the old downbeat was.
FOUT is useless for Flex loops as fade out is added after rec end.
Yes, that’s true. But it is used with Pickup machines for this purpose.
I don’t really like Pickup machines, but they were (I guess) how Elektron tried to satisfy the sound-on-sound type loopers.
Thx @Bryan_T for the effort! Interesting findings.
That’s why I usually sequence everything from my OT (usually no clock, just notes with tempo manually matched). I even sometimes drag samples/recordings that were made with software only or were recorded from another instrument (often rendered and reimported into my daw) into a sampler plugin in my daw, trigger the playback from my OT and record into OT instead of transferring via usb.
Longer samples I record with manual quantized recording start/stop into OT.
If software/daw is also clocked from OT, i give it a few bars until the sampling process is started to get a stable clock.
2 Likes
No details about FIN / FOUT in the manual. Their curve is constant. If you increase their value, it adds silence without changing the curve.
Very bad fades…
1 Like
I realized that there are more options on the table if you set RLEN to “MAX” and then scale the track timing with the 1/8, 1/4, 1/2, 3/4, 1, 3/2, or 2 values. Everything I posted previously only used 1. I’ve uploaded a new spreadsheet below. In addition to the BPM (Octatrack’s displayed value) and number of trigs, I’ve added a column for the clock division and also calculated the total length of the loop.
I think the longest clickless quantized loop that you can get out of the Octatrack is 250.7755 seconds. Not bad!
octatrack_clickless.xlsx (66.8 KB)
2 Likes
Can be doubled imo (508s being max with empty RAM)
I’m not sure how you’d do that that would guarantee it was clickless and quantized. The 250.7755 seconds is 64 trigs, 1/8 multiplier, with a BPM of 30.6. How would you double it?